Anonymous posted on Wednesday, June 22, 2005 - 5:45 pm
1. I find that if I specify MISSING ARE ALL(999); in my Mplus script files, any case with any variable with a value of 999 is deleted from the analysis; even if the variable with a 999 value is one that is not used in the model (i.e., is not included on the USEVARIABLES ARE statement). Is this how MISSING ARE should work ?
2. I find that when I use MISSING ARE ALL(999) and TYPE=TWOLEVEL RANDOM MISSING when my outcome variables include a categorical and ordinal variable (i.e., both are simultaneous outcomes with BETWEEN-CLUSTER mean of variation), Mplus omits entire clusters from analysis if only one of the cases in the cluster has variable with a value of 999. Is this how MISSING ARE should work ?
1. This is not how it works. Listwise deletion applies only to the variables on the USEVARIABLEs list. If this is not clear, please send your output, data, and license number to firstname.lastname@example.org and we can show you what is happening.
2. You should also send this to support. I have never seen this happen. You may be misunderstanding something.